Nerve growth factor (NGF) was the first neurotrophin to be identified, and its role in the development and survival of both peripheral and central neurons has been well characterized. NGF has been shown to be a critical survival and maintenance factor in the development of peripheral sympathetic and embryonic sensory neurons and of basal forebrain cholinergic neurons (Smeyne, et al., Nature 368:246-249 (1994); Crowley, et al., Cell 76:1001-1011 (1994)). NGF upregulates expression of neuropeptides in sensory neurons (Lindsay, et al., Nature 337:362-364 (1989)), and its activity is mediated through two different membrane-bound receptors, the TrkA tyrosine kinase receptor and the p75 receptor which is structurally related to other members of the tumor necrosis factor receptor family (Chao, et al., Science 232:518-521 (1986)).
In addition to its effects in the nervous system, NGF has been increasingly implicated in processes outside of the nervous system. For example, NGF has been shown to enhance vascular permeability (Otten, et al., Eur J. Pharmacol. 106:199-201 (1984)), enhance T- and B-cell immune responses (Otten, et al., Proc. Natl. Acad. Sci. USA 86:10059-10063 (1989)), induce lymphocyte differentiation and mast cell proliferation and cause the release of soluble biological signals from mast cells (Matsuda, et al., Proc. Natl. Acad. Sci. USA 85:6508-6512 (1988); Pearce, et al., J. Physiol. 372:379-393 (1986); Bischoff, et al., Blood 79:2662-2669 (1992); Horigome, et al., J. Biol. Chem. 268:14881-14887 (1993)). Although exogenously added NGF has been shown to be capable of having all of these effects, it is important to note that it has only rarely been shown that endogenous NGF is important in any of these processes in vivo (Torcia, et al., Cell. 85(3):345-56 (1996)). Therefore, it is not clear what the effect might be, if any, of inhibiting the bioactivity of endogenous NGF.
NGF is produced by a number of cell types including mast cells (Leon, et al., Proc. Natl. Acad. Sci. USA 91:3739-3743 (1994)), B-lymphocytes (Torcia, et al., Cell 85:345-356 (1996), keratinocytes (Di Marco, et al., J. Biol. Chem. 268:22838-22846)), smooth muscle cells (Ueyama, et al., J. Hypertens. 11: 1061-1065 (1993)), fibroblasts (Lindholm, et al., Eur. J. Neurosci. 2:795-801 (1990)), bronchial epithelial cells (Kassel, et al., Clin, Exp. Allergy 31:1432-40 (2001)), renal mesangial cells (Steiner, et al., Am. J. Physiol. 261:F792-798 (1991)) and skeletal muscle myotubes (Schwartz, et al., J Photochem, Photobiol. B 66:195-200 (2002)). NGF receptors have been found on a variety of cell types outside of the nervous system. For example, TrkA has been found on human monocytes, T- and B-lymphocytes and mast cells.
An association between increased NGF levels and a variety of inflammatory conditions has been observed in human patients as well as in several animal models. These include systemic lupus erythematosus (Bracci-Laudiero, et al., Neuroreport 4:563-565 (1993)), multiple sclerosis (Bracci-Laudiero, et al., Neurosci. Lett. 147:9-12 (1992)), psoriasis (Raychaudhuri, et al., Acta Derm. l'enereol. 78:84-86 (1998)), arthritis (Falcimi, et al., Ann. Rheum. Dis. 55:745-748 (1996)), interstitial cystitis (Okragly, et al., J. Urology 161:438-441 (1991)) and asthma (Braun, et al., Eur. J Immunol. 28:3240-3251 (1998)).
Consistently, an elevated level of NGF in peripheral tissues is associated with inflammation and has been observed in a number of forms of arthritis. The synovium of patients affected by rheumatoid arthritis expresses high levels of NGF while in non-inflamed synovium NGF has been reported to be undetectable (Aloe, et al., Arch. Rheum. 35:351-355 (1992)). Similar results were seen in rats with experimentally induced rheumatoid arthritis (Aloe, et al., Clin. Exp. Rheumatol. 10:203-204 (1992)). Elevated levels of NGF have been reported in transgenic arthritic mice along with an increase in the number of mast cells. (Aloe, et al., Int. J. Tissue Reactions—Exp. Clin. Aspects 15:139-143 (1993)).
Treatment with exogenous NGF leads to an increase in pain and pain sensitivity. This is illustrated by the fact that injection of NGF leads to a significant increase in pain and pain sensitivity in both animal models (Amann, et al., Pain 64, 323-329 (1996); Andreev, et al., Pain 63, 109-115 (1995)) and human (Dyck, et al., Neurology 48, 501-505 (1997); Petty, et al., Annals Neurol. 36, 244-246 (1994)). NGF appears to act by multiple mechanisms including inducing the neurotrophin BDNF (Apfel, et al., Mol. Cell. Neurosci. 7(2), 134-142 (1996); Michael, et al., J Neurosci 17, 8476-8490 (1997)) which in turn changes pain signal processing in the spinal cord (Hains, et al., Neurosci Lett. 320(3), 125-8 (2002); Miletic, et al., Neurosci Lett. 319(3), 137-40 (2002); Thompson, et al., Proc Natl Acad Sci USA 96(14), 7714-8 (1999)), inducing changes in the peripheral and central connections of the sensory neurons and other pain-transmitting neurons in the spinal cord (Lewin, et al., European Journal of Neuroscience 6, 1903-1912 (1994); Thompson, et al., Pain 62, 219-231 (1995)), inducing changes in axonal growth (Lindsay, R M, J Neurosci. 8(7), 2394-405 (1988)) inducing bradykinin receptor expression (Peterson et al., Neuroscience 83:161-168 (1998)), inducing changes in expression of genes responsible for nerve activation and conduction such as ion channels (Boettger, et al., Brain 125(Pt 2), 252-63 (2002); Kerr, et al., Neuroreport 12(14), 3077-8 (2001); Gould, et al., Brain Res 854(1-2), 19-29 (2000)), potentiating the pain related receptor VR1 (Chuang, et al., Nature 411 (6840), 957-62 (2001); and causing pathological changes in muscles (Foster, et al., J Pathol 197(2), 245-55 (2002)). Many of these changes take place directly on the pain transmitting sensory neurons and apparently are not dependent on concomitant inflammation. In addition, there are at least two other cell types known to respond to NGF and that may be involved in changes of pain sensation or sensitivity. The first of these, the mast cell, has been reported to respond to NGF with degranulation (Yan, et al., Clin. Sci. (Lond) 80:565-569 (1991)) or, in other studies, to cause or increase mediator production or release in collaboration with other agents (Pearce and Thompson, J. Physiol. 372:379-393 (1986), Kawamoto, et al., J. Immunol. 168:6412-6419 (2002)). It has clearly been shown in the rat that NGF mediated pain responses are at least somewhat mediated by mast cells (Lewin, et al., Eur. J. Neurosci. 6:1903-1912 (1994), Woolf, et al., J. Neurosci. 16:2716-2723 (1996) although the potential relevance of this remains to be shown in humans. Primary sympathetic neurons are also known to respond to NGF and to also be involved in pain signaling (Aley, et al., Neuroscience 71:1083-1090 (1996)). It is clear that removing sympathetic innervation modifies the hyperalgesia normally seen in response to treatment with NGF (Woolf, et al., J. Neurosci. 16:2716-2723 (1996)).
Twenty-three million patients have surgical procedures each year. Pain is usually localized within the vicinity of the surgical site. Post-surgical pain can have two clinically important aspects, namely resting pain, or pain that occurs when the patient is not moving and mechanical pain which is exacerbated by movement (coughing/sneezing, getting out of bed, physiotherapy, etc.). The major problem with post-surgical pain management for major surgery is that the drugs currently used have a variety of prominent side effects that delay recovery, prolong hospitalization and subject certain vulnerable patient groups to the risk of serious complications. Post-surgical pain, or pain that occurs after surgery or traumatic injury is a serious and often intractable medical problem.
There are two general categories of medication for the treatment of pain, both of which have disadvantages. The first category includes the nonsteroidal anti-inflammatory drugs (NSAIDs) which are used to treat mild or moderate pain, but whose therapeutic use is limited by undesirable gastrointestinal effects such as gastric erosion, the formation of peptic ulcer or the inflammation of the duodenum and of the colon. NSAIDs also can cause renal toxicity with prolonged use, and further, as described below, are not very effective for treating pain associated with or arising from certain conditions, including post-surgical pain. The second category includes morphine and related opioids which are used to treat moderate to severe pain but whose therapeutic use is limited because of undesirable effects such as sedation, confusion, constipation, respiratory depression, renal colic, tolerance to prolonged use and the risk of addiction. Compounds useful for treating pain with fewer or no side effects are therefore needed.
Pain is often categorized as “inflammatory”, “neuropathic” or “visceral”, but these traditional general labels have inherent problems. They imply mechanistic similarity or identity among all sources of pain within one of these very general categories. In fact, there are many different types of inflammatory pain and sources of pain that are neither inflammatory nor neuropathic. Further, types of pain that have an inflammatory component, and/or are traditionally termed “inflammatory”, does not mean that other physiological aspects do not contribute to the pain state. For example, both osteoarthritis and interstitial cystitis would be defined by their names as sterile inflammatory conditions of respectively joints or the urinary bladder, but it is clear that the pains associated with these two conditions are mechanistically quite different from each other. This is indicated by the varying effects of a given type of anti-pain medication with respect to these types of pain. The majority of patients with osteoarthritis receive good pain relief (at least initially) with NSAIDs. However, NSAIDs treatment is completely ineffective with interstitial cystitis.
Post-surgical pain (interchangeably termed, post-incisional pain) is often considered a variety of inflammatory pain. While there may be an “inflammatory” component to post-surgical pain, clearly additional mechanisms are involved. For example, during surgery or other injury, both vasculature and nerves are cut or torn. This does not happen in a tissue undergoing only inflammation. It is clear that cutting a nerve can induce ongoing activity, which is perceived as painful. In addition, severing blood vessels lead to a tissue that is relatively ischemic, also a painful stimulus that is not present during inflammation alone.
The different mechanisms involved in surgical or injury-induced pain as compared to inflammation is exemplified by the varying pharmacology and underlying anatomical substrates of pain relief in the two conditions. Yamamoto, et al., (Brian Res. 909(1-2):138-144 (2001)) have shown that inhibition of spinal N-acetyl-alpha-linked acidic dipeptidase (NAALADase) causes a marked attenuation of mechanical pain which accompanies the inflammatory stimulus of carrageenan injection. However, in parallel experiments where NAALADase was inhibited in an identical fashion after an incision, there was no attenuation of mechanical pain. These observations demonstrate that the biochemistry or pharmacology underlying post-surgical pain is distinct from those underlying inflammatory pain. The anatomical structures important in modulating pain sensation have also been examined in post-surgical and other pain states (Pogatzki, et al., Anesthesiology, 96(5):1153-1160 (May 2002)). Descending influences for the brainstem, more specifically the rostral medial medulla, are important modulators of secondary hyperalgesia in general inflammatory, neuropathic and visceral pain states. When the brain stem area was lesioned, no change in any pain response measured after incision was observed. These results indicate that primary and secondary hyperalgesia after an incision are not modulated by descending influence from the RMM. The lack of contribution of descending facilitatory influences from the RMM to secondary hyperalgesia after gastrocnemius incision supports the notion that incision-induced pain involved dissimilar mechanisms compared with inflammatory and neuropathic pain. In addition to the obvious differences in post-surgical or injury-induced pain from inflammatory, visceral or neuropathic pain, these results demonstrate that the mechanisms involved in post-surgical pain (or injury-induced pain) are clearly different from other pains. Further, the utility of a particular pharmacological (or other) intervention in treating post-surgical pain is not predictable by testing that pharmacological agent or intervention in inflammatory, visceral or neuropathic pain models.
Disappearance of pain at rest and persistence of pain with activities and in response to mechanical stimuli at the wound site is also present in patients after surgery. (Moiniche, et al., Acta Anaesthesiol. Scand. 41:785-9 (1997)). Studies suggest that pain at rest and evoked pain caused by incisions are likely transmitted by different afferent fiber populations and/or different receptors. Other than using local anesthetics to inhibit these evoked responses, few drugs that markedly reduce pain with coughing and movement after surgery are available.
Pretreatment with a local anesthetic to block the pain during the experimental incision has been shown to initially prevent ongoing pain and the primary mechanical hyperalgesia. Pain from the incisions also disappears when lidocaine is injected after the injury. However, as the local anesthetic effect abates, the primary hyperalgesia returns. In patients, local anesthetic injections made before surgery are roughly equivalent for reducing pain to injections made after surgery. (Moiniche, et al., Anesthesiology 96:725-41 (2002))
Clinical studies experiments in human volunteers, and a preclinical incision model agree that administration of local anesthetic before or after the incision are roughly equivalent. The activation of central pain transmitting neurons during incision and sensitization are not necessary for pain behaviors several days later. Rather, for incisions, enhanced responsiveness of central neurons and pain require ongoing afferent input from the incision. After any preincision analgesic treatment abates, the surgical wound appears capable of reinitiating sensitization and regenerating the pain responses. (Pogatzki, et al., J Neurophysiol 87:721 (2002))
The area of hyperalgesia (including the uninjured zone) caused by incisions has also been mapped. Secondary hyperalgesia (hyperalgesia outside the injured area) is one measure of enhanced responsiveness of the central nervous system, i.e. central sensitization. It has been noted that the area of flare or redness (possibly a result of axon reflexes) caused by incision was distinct from the area of hyperalgesia. As opposed to pain at rest and primary mechanical hyperalgesia, the large area of hyperalgesia never developed when local anesthetic injection was made before the incision. Moreover, it could not be reversed by local anesthetic injection after incision. In patients after surgery, in some cases, certain treatments greatly reduce the area of hyperalgesia but do not greatly modify clinical measures of post-surgical pain (pain scores and opioid consumption). It has been shown that reducing the area of hyperalgesia after colectomy did not greatly reduce acute pain but was associated with a decrease in the number of patients that developed residual pain even as late as 6 months after colectomy. (De Kock, et al., Pain 92:373-80 (2001)).
The use of anti-NGF antibody to treat chronic visceral pain has been described. See PCT Publication No. WO 01/78698. Brennan et al. report administration of TrkA immunoadhesin in a rat model of post-surgical pain. See Society for Neuroscience Abstracts 24(1-2) 880 (1998).
All references cited herein, including patent applications and publications, are incorporated by reference in their entirety.